Linking governance with environmental quality: a global perspective

Sustainable environmental quality is a global concern, and a concrete remedy to overcome this challenge is a policy priority. Therefore, this study delves into the subject and examines the effects of governance on environmental quality in 180 countries from 1999 to 2021. To maintain comparability and precision, we first classify countries into full and income-level panels and then, innovatively, construct a composite governance index (CGI) to capture the extensive effects of governance on CO2 emissions. Complementing the stationarity properties of the variables, we employ the cross-sectionally augmented autoregressive distributed lags model to analyze the data. Our survey yields four key findings. First, a long-run nexus between CGI, CO2 emissions, and other control variables is confirmed. Second, the findings indicate that CGI is crucial to improving environmental quality by reducing CO2 emissions across all panels. Third, we find that while CGI maintains a similar magnitude, the size of its effects substantially varies according to the income level of the underlying countries. Fourth, the findings reveal that energy consumption, population growth rate, trade openness, and urbanization contribute to environmental degradation, while financial development and the human development index are significant in reducing CO2 emissions. Our findings suggest specific policy implications, summing up that one common policy is not a good fit for all environmental quality measures.


Literature review
Good governance is a complex and multidimensional process of evaluating the extent to which public institutions manage the available resources, perform institutional affairs, and ensure that human rights are realized in a way that is essentially free of fraud and corruption with due consideration for the rule of law 29 .Good governance ensures that a nation's interests are protected through effective conduits for governing and managing existing and potential resources 30 .North 31 , Greif 32 , and Acemoglu et al. 33 promoted the concept of governance through conduits of economic, social, judicial, and political elements that highly impact macro-level policies to preserve public resources for significant social inclusion, prosperity, and the wellbeing of a nation.Theories predict that good governance plays an essential role in the formulation of policies and practices that ensure a participatory development viewpoint through increasing people's agency in the sense of process freedom concerning environmental policies.This means allowing both governments and individuals to actively engage in, plan for, and implement policies based on their development priorities and needs 34 .Numerous studies have examined the impact of good governance on a number of socioeconomic indicators such as growth, finance, health outcomes, food insecurity, and poverty across various geographical contexts [34][35][36][37] .However, the effects of good governance on environmental degradation have not been extensively studied, but there are some studies worth reviewing.For instance, Shabir et al. 38 investigated the effects of governance, innovative technologies, trade openness, and economic growth on CO 2 emissions in a panel of Asia-Pacific Economic Cooperation (APEC) member countries over the period from 2004 to 2018, using the common correlated effects mean group technique.The authors observed a bidirectional link between governance and CO 2 emissions.Wang et al. 39 explored the asymmetric effects of institutional quality, environmental governance, and technological innovations on ecological footprints.They employed a set of panel data for European Union countries from 1990 to 2019 and a series of dynamic panel regression methods.They noticed that innovation, institutional quality, and environmental governance are crucial to reducing the ecological footprint across the reviewed countries.Sibanda et al. 28 examined the effects of governance on natural resources and environmental degradation from 1994 to 2020 using the generalized method of moments (GMM) technique.Their findings lend support for a statistical association between governance and environmental degradation.They also found that the rapid environmental degradation is significantly caused by the reluctance of the government to implement rules and regulations in the region.Xaisongkham and Liu 40 delved into the effects of governance on environmental degradation in a set of selected developing economies from 2002 to 2016.The authors employed the GMM technique and found that the rule of law and government effectiveness are significant factors in reducing environmental degradation in developing countries.They suggested that sustainable environmental quality entails effective institutions to regulate human behavior with respect to environmental protection.In the same vein, Jahanger et al. 41 used autocracy and democracy as proxies for governance quality and examined their effects on CO 2 emissions in a panel of 69 developing countries over the period from 1990 to 2018.The authors employed panel cointegration and FMOLS methods and confirmed that governance quality has a long-run relationship with CO 2 emissions.They also confirmed that democracy significantly reduces environmental pressures, while globalization and financial development impose adverse effects on the environment.
The literature also reveals that Azam et al. 42 evaluated the impact of good governance on environmental quality and energy consumption in a panel of 66 developing countries for the period spanning from 1991 to 2017 using the GMM method.The authors constructed a governance index using three indicators such as political stability, administrative capacity, and democratic accountability.They observed that, though good governance has been significantly positive in affecting energy consumption, globalization has been found to be insignificant in increasing environmental quality.Moverover, Gök and Sodhi 43 examined the link between governance and environmental quality in a panel of 115 countries classified as high-, middle-, and low-income countries from 2000 to 2015.The authors employed the system-GMM model and noticed that good governance improves environmental quality in high-income countries while having an adverse effect in middle-and low-income countries.Their conclusions suggested that improving the quality of governance is essential to environmental outcomes without tampering with existing policies.Contrary to this, Udemba 44 investigated the effects of good governance on environmental quality in Chile using a set of time-series data from the first quarter of 1996 to the fourth quarter of 2018 and a non-linear regression approach.The author found that both good governance and foreign direct investments are statistically significant for improving environmental quality in Chile.Furthermore, Ahmed et al. 45 examined the asymmetric effects of good governance, financial development, and trade openness on environmental degradation in Pakistan over the period from 1996 to 2018.The authors employed autoregressive distributive lags (ARDL) and non-linear ARDL models to test their hypotheses.In addition to confirming a long-run nexus between the predictors, the authors found that positive shocks to financial development and institutional quality have a significant effect on environmental degradation, while the quality of institutions is highly sensitive to enhancing environmental quality.
Akhbari and Nejati 46 proxied governance by corruption index in a panel of 61 developing countries from 2003 to 2016 using a dynamic panel threshold model.They observed that an increase in the corruption index above a certain threshold level causes environmental quality to decrease in developing countries while having an insignificant impact below the threshold level.Dhrifi 47 also assessed the impact of governance on environmental degradation in a panel of 45 African countries over the period 1995 to 2015 using the GMM technique.The author noticed a positive relationship between governance and environmental degradation and a negative link with health outcomes.Further, Wawrzyniak and Doryń 48 investigated the influence of good governance on moderating the relationships between economic growth and CO 2 emissions in a panel of 93 emerging and developing economies from 1995 to 2014.The authors used government effectiveness and control of corruption indicators as proxies for governance and employed the GMM model.Their findings revealed that government effectiveness is significant in moderating the influence of economic growth on CO2 emissions.Similarly, Samimi et al. 49 employed a set of annually aggregated datasets for a panel of 21 countries in the Middle East and North Africa from 2002 to 2007 to examine the impact of good governance on environmental degradation.The authors used three indicators, such as government effectiveness, regulatory quality, and control of corruption, as proxies for good governance.They found that government effectiveness has a positive effect on environmental quality, while the remaining two indicators were found to be insignificant.Finally, Tamazian and Rao 50 investigated the relationships between financial development, environmental degradation, and good governance in a panel of 24 transitional economies from 1993 to 2004.Using the standard reduced-form modeling approach and GMM models, the authors found that both financial development and good governance (institutional quality) are crucial factors for environmental performance.
Recent studies have significantly contributed to enhancing the contemporary body of knowledge in the field; however, a critical review of the cited studies reveals several gaps.First, good governance is a multifaceted concept, and its precise effects may not be well examined by using single or inconclusive proxies.For example, www.nature.com/scientificreports/various studies employed different proxies for good governance, among which government effectiveness and control of corruption are the most common ones.To rectify this issue, we developed the following hypothesis: Hypothesis 1: Composite governance index (CGI) is an accurate predictor that allows more precise evaluation of the effects of good governance on the subject.
Second, prior studies achieved conflicting results about the effects of good governance on environmental quality, leaving the subject unattended to offer specific policy implications.Therefore, to address this empirical shortcoming, the following hypothesis is developed: Hypothesis 2: CGI has a long-term and positive link with CO 2 emissions.
Third, the review of recent studies reveal that holistic measures to highlight global perspectives and precise comparability of the effects of good governance on environmental quality are missing.To address this empirical shortcoming, we developed the following hypothesis: Hypothesis 3: Based on the size of the underlying economies, the effect size of good governance varies and thus exhibits non-monotonic behavior.

Methodology
In this section, we explain the methodological approach used in the study to assess the effects of good governance on CO 2 emissions.This approach has been widely used in prior literature and leads to a systematic way of testing the hypotheses developed 51,52 .Although we describe the methods sequentially in the following sub-sections, we summarize them through a visual abstract depicted in Fig. 2.

Data presentation.
The present study focuses on the effects of good governance on environmental degradation in 180 countries from 1999 to the most recent updated datasets in 2021.Table 1 presents the list of reviewed countries.Based on the primary objective of the study, we first group the countries into a full panel and then into income level categories such as high-income (HIC), upper-middle-income (UMIC), lower-middleincome (LMIC), and low-income (LIC).The classification is based on the World Bank's 53 report and allows us to maintain rational comparability of the results to offer a global image of the nexus between good governance and environmental degradation.
Selection and description of variables.We use a set of variables that are consistent with the theoretical framework and recent empirical works (see, for instance, [54][55][56] ), except for the CGI, which is innovatively   58 and six governance indicators such as control of corruption, government effectiveness, political stability, the rule of law, regulatory quality, and voice and accountability.For two reasons, it is important to construct a CGI.First, it is a more efficient approach to exploring the extensive effects of good governance on the subject compared to individual indicators and other index construction methods.Second, the incorporation of CGI allows the study to include more control predictors, leading to an appropriate specification and more accurate results [59][60][61] .Table A1 of Appendix A explains CGI's construction process in detail.CGI is expressed in numbers ranging from 0 (imperfect) to 1 (perfect) governance.
Measurement of income level.GDP growth rate (EG) has been used to present economic variations through various stages of development at which CO 2 emissions are produced 62 .EG is expressed as an annual percentage.
Measurement of financial development.The financial development index (FDI) of the International Monetary Fund has been used as the best-fit proxy for financial development.FDI is expressed in numbers from 0 to 1 (high).Recent studies indicate that financial development influences CO 2 emissions 63,64 .Therefore, we control for the effects of FDI on CO 2 emissions.
Measurement of energy consumption.Energy consumption (EGY), expressed in kilograms of oil equivalent per capita, is used as a control variable.Recent studies suggest the use of EGY as a key pollutant predictor in the analysis of environmental quality and other socioeconomic indicators.It is evident that EGY supports higher growth 65 , while it also increases the use of fossil fuels, resulting in higher CO 2 emissions.Chontanawat 66 and Elfaki et al. 67 argue that there is a triangle causal link between EGY, EG, and CO 2 emissions.
Measurement of human interaction.In order to control for the effects of human interaction on CO2 emissions, we employ three common variables, namely, the human development index (HDI), population growth rate (PGR), and urbanization (URB).HDI, PGR, and URB, respectively, are expressed in numbers from 0 to 1 (high), annual growth rate, and percentage of the total population.Studies indicate that human intervention has substantively disturbed the contemporary ecosystem.However, effective administration of human activities, as well as utilizing their potential, may improve environmental quality 21,68 .Moreover, a higher proportion of greenhouse gas emissions is linked to the process of global urbanization, which is primarily evident in nations following growth-targeting regimes 56 .These emissions are mostly produced by construction projects, higher energy consumption, and the use of chemical materials.
Measurement of trade openness.Trade openness (TOP), expressed as a percentage of GDP, is our final control variable.Though recent literature is largely inconclusive about the effects of TOP on CO 2 emissions 69 , two main findings-positive and negative impacts-are evident.The study incorporates TOP into the analysis to avoid any potential spuriousness.

Sources of data.
The datasets relevant to governance indicators come from Worldwide Governance Indicators (WGI).The datasets for FDI have been collected from the International Monetary Fund (IMF), while the required datasets for HDI were compiled from PWT 9.0 (Penn World Table ), sourced from Feenstra et al. 70 .The data for all other variables has been collected from World Development Indicators (WDI).

Model specification.
Our main primary objective is to examine the effects of CGI-that is, the composite governance index-on CO 2 emissions in a large panel to represent a global image.Assuming that good governance is essential to environmental quality, as suggested by theoretical expectations of institutional impacts 71 , we initiate with the following dynamic panel multivariate specification: where all variables are defined before, η i = intercept, 1i , ..., 8i = long-run coefficients, and n t = country-specific unobserved effects.The estimation of Eq. (1) requires us to select and compute a number of econometric techniques that are explained in the following sub-sections.

Cross-sectional dependence test.
In panel data analysis, appropriate specification requires several prior estimations, one of which, in particular, is the cross-sectional dependence (CD) test.Rapid globalization, unrestricted trade, common technological deployment, and capital mobility are some obvious reasons why countries may exhibit CD 72 .Thus, we begin with the CD test of Pesaran 73 , which takes the following form: where ⌢ ρ ij is the sample estimates of the pair-wise correlation of the residuals and " T ij = # (T i ∩ T j ) is the number of common time-series observations between unit i and j. " Eq. (2) shows that under the null hypothesis of no cross-sectional dependence CD d − → N(0, 1) for N → ∞ and T 74 .To ensure the robustness of the results obtained (1) from Eq. ( 2), we use the proposed model of Pesaran and Yamagata 75 to tes the null of slope homogeneity of the panels under review.

Stationarity test.
Next, in light of the rejected null of no CD, the common panel unit root test may generate inconsistent results that may lead to misspecification.Therefore, we use the proposed test of Pesaran 76 , the socalled CIPS (cross-sectionally augmented Im, Pesaran, and Shin) method.It is based on the foundational crosssectionally augmented Dickey and Fuller (CADF) test with augmented cross-sectional mean y it and differenced cross-sectional mean value y it of the variables under review as follows: where as a common test of the null γ = 0 for every i against its alternative γ i < 0, ..., γ N0 < 0, N 0 ≤ N and then, given by the average of individual CADF as: As a common practice, it notes that for the rejected null of panel non-stationarity, the critical value of a desired significant level must be less than the CIPS test statistics at the level.CIPS is advantageous over other panel unit root tests.It neatly detects the true stationarity of the panel variables arising from common unobserved factors 76 , thus leading to an appropriate specification.
Cointegration tests.Again, for the rejected null of no CD, common panel cointegration techniques may be biased.Thus, we employ the proposed model of the Westerlund 77 test, which has two key advantages over other panel cointegration methods.First, it accounts for the effects of any CD existing in the panel, and second, it considers the lead-lag length for small samples.The study employs the following compact form of the test: where d t (1, t) ′ = deterministic regressor, η ′ i = vector of parameters (η 1i , η 2i ) ′ , and other parameters hold similar meaning as explained before.To estimate the error-corrected form through the least squares method, we modeify Eq. ( 5) and represent it as follows: Having all vectors defined before,ϕ i = speed of adjustment at which the model returns to its initial equilib- rium.Moreover, Eq. ( 6) adjusts the errors to be independent across all t and i.It also corrects for any CD through bootstrapping method.

CS-ARDL model.
To examine the effects of CGI and other control variables on CO 2 emissions in a group of panels, we use the CS-ARDL (cross-sectionally augmented autoregressive distributed lags) model of Chudik and Pesaran 78 , which is an appropriate technique for the case of our inquiry.The rationality of using the CS-ARDL model is based on two key empirical reasons.First, for the rejected null of no CD, common panel techniques fail to capture the true effects and may produce inconsistent and biased coefficients.Second, it corrects any slope heterogeneity and allows the variables to exhibit mixed stationarity properties.Having said that, we proceed to specify the CS-ARDL model by augmenting the symmetric ARDL with a linear combination of cross-sectional mean values of the lagged dependent variable and explanatory variables to capture the CD in the error term as follows: where y it = dependent variable, x it = explanatory variables, u = lag operator of the dependent variable, v = lag operator of the independent variables, ϑ i = intercept, i = speed of adjustment of panel to its long-run equilib- rium, (y it−1 − η i ′x it−1 + ϕ −1 i y t + ϕ −1 i ξ ′ x t ) = level terms of CD and cointegration between variables, y = cross- sectional mean of the dependent variable, x = cross-sectional mean of the explanatory variables, i , η i , and ϕ i = long-run coefficients, γ i , φ i, ξ i , and ̟ i = short-run coefficients, and u it = the error-term.Equation ( 6) uses either mean group or pooled mean group estimators, the choice of which depends on the homogeneity of the slope coefficients of the long-term effects 79 .In our dynamic panel estimations, the study employs the pooled mean group estimator, observing it as an appropriate method to preserve the consistency and efficiency of coefficients.

FMOLS and DOLS tests.
As suggested by prior literature (refer 38,80,81 ), we test the robustness of our estimated outcomes obtained from the computation of Eq. ( 7) using the fully modified least squares (FMOLS) and (3) (5) dynamic ordinary least squares (DOLS) techniques.Phillips and Hansen 82 developed the FMOLS model to estimate an optimal cointegrating equation; however, based on the preference for correcting serial correlation and endogeneity bias, we apply the FMOLS method proposed by Pedroni 83 as expressed below: where ) and ⌢ L i presents the lower trian- gulation of ⌢ � i .Moreover, DOLS is a rather parametric technique and has a similar asymptotic distribution to that of FMOLS 84 .We use and report both to confirm the consistency and robustness of the estimated outcomes.

Results and discussions
Descriptive summary.A summary of statistics has been provided to reflect the overall state of the predictors used in the study.While Table 2 presents the summary statistics for the full panel, it also disaggregates them by income-level groups.Though one can read through it, it shows that the mean values of CO 2 emissions and CGI, two variables of interest, are 5.091 metric tons per capita and 0.746, respectively, in the full panel.Better insight is achieved when the statistics are compared by income level.It shows that despite the fact that HIC has the highest mean value of the CGI, it also produces the highest CO 2 emissions compared to UMIC, LMIC, and LIMC.
Using actual series, we further generate an annual average trend of the CGI and CO 2 emissions across income groups and depict them in Figs. 3 and 4, respectively.
Figure 3 indicates that CGI has been smoothly improving over the years in HIC and LIC, while LMIC and UMIC panels exhibit some structural breaks.On the other hand, CO 2 emissions were significantly reduced over the years in HIC and LIC, whereas in UMIC and LMIC, a downward shift was only evident from 2018 onwards (Fig. 4).Furthermore, before any empirical analysis, we estimated the pairwise correlation matrix and found (8) www.nature.com/scientificreports/ that there is no significant correlation between all the variables, both in the full and income-level groups.To save space, we avoided reporting the correlation analysis, but it can be provided upon request.

CD and unit root test results.
Prior to any inferences, as described earlier, the study tests the null of no CD among the full and the income-level panels.The results are reported in Table 3 and indicate that except for URB in full, UMIC, and LMIC panels, all other variables are significant to reject the null of no CD.Moreover, the study examines the slope heterogeneity using the proposed model of Pesaran and Yamagata 85 .It is also used to ensure that the results derived from the CD test are consistent.The results reported in Tables 4 and 5 indicate that for all panels, the null is rejected at 1% and 5% significant levels, implying that the slopes are heterogeneous across the panels.The results obtained from the CD test suggest examining the unit root of the variables.To this end, we use the CIPS test of Pesaran 76 and report the results in Table 7.The results demonstrate that EG, FDI, and HDI are significant to reject the null of the unit root at the level, while the remaining variables are first-difference stationary in the full and LMIC panels.For HIC, only EG is level-stationary, and the other predictors are first-differenced stationary.The results for the UMIC panel, FDI, and HDI reject the null, while others become significant to    102 , and Jahanger et al. 41 , who employed a single dimension and found that governance (institutional quality) has a negative impact on CO 2 emissions.Nevertheless, our findings fully support the outcome of studies conducted by Shabir et al. 38 , Wang et al. 39 , Sibanda et al. 28 , and Xaisongkham and Liu 40 .
In terms of the control variables, the findings show that EGY has a significant impact on CO 2 emissions both in the short-and long-run across all panels.It indicates that in the short (long) run, a unit increase in EGY causes CO 2 to increase by 0.328 (0.741), 0.717 (0.312), 0.522 (0.227), 0.191 (113), and 0.229 (0.100) metric tons per capita in the full, HIC, UMIC, LMIC, and LIC panels, respectively.The results imply that higher EGY produces more CO 2 emissions.These findings show that EGY is yet another vital component that directly leads to the degradation of environmental quality worldwide.Due to rapid development, energy demand has been increasing around the world 103 .The burning of fossil fuels is used to meet a sizable percentage of this need.As a result, energy use significantly contributes to the decrease in environmental quality.Our results support the findings of Javid and Sharif 104 for Pakistan; Shahbaz et al. 105 for low, middle, and high-income countries; Farhani and Ozturk 106 for Tunisia; Beşe and Kalayci 107 for Egypt, Kenya, and Turkey; and Adebayo and Kirikkaleli 108 for Japan; but contrast with those of Jebli et al. 109 for OECD member countries; and Shafiei and Salim 110 , who, respectively, provided significant statistical evidence that more EGY has a reversible effect on CO 2 emissions.
Except for the high-income panel, the coefficient of EG yields a positive sign at 10% significance, implying that EG accelerates CO 2 emissions in UMIC, LMIC, LIC, and the full panel.Specifically, a 1% increase in EG causes CO 2 emissions to increase by 0.021 (0.213), 0.047 (0.039), 0.061 (0.124), and 0.023 (0.201) metric tons per capita in the short (long) run, respectively, in the full, upper-middle-income, lower-middle-income, and low-income panels.These results correspond to those of Pilatowska et al. 111 for the EU, Kasman and Duman 112 for new EU members, Bekun et al. 113 for 16 EU members, Saidi and Rahman 114 for OPEC countries, and Khan 115 for South Asian economies.The results are linked to stylized facts.It is expected that environmental quality will pay a price with an increase in overall economic output and national consumption.This implies that when the use of non-renewable resources increases, environmental degradation also increases, and thus, the potential loss of environmental ecosystems is only one of the negative effects of rapid economic growth on the environment.However, not all types of growth harm the environment.A sound allocation of funds to environmental preservation when real earnings rise is found to be effective and, as such, good governance.
Altogether, financial development, as proxied by the financial development index (FDI), negatively affects CO 2 emissions.Literally, it was expected that a well-developed financial sector would facilitate enhanced access to higher investments in lower carbon emission production that significantly decreased CO 2 .Magazzino 116 also found that financial development has a negative impact on CO 2 emissions.Further studies by Al-Mulali et al. 117 , Tang and Tan 118 , Ho and Ho 119 , and Rahman and Alam 20 also emphasize that a well-developed financial sector and access to credit significantly reduce CO 2 emissions due to informed and well-thought-out investments in low-carbon-producing projects.In the purview of human interaction with the environment, we regressed HDI on CO 2 emissions and found that, in contrast to a vast number of prior studies, HDI is significant for reducing CO 2 emissions in the long run.This might be due to the selection of proxies.Before augmenting HDI, we regressed HCI (human capital index) and found that it has a rather positive impact on CO 2 emissions.While HCI does not fully cover all aspects of human interaction with society, we swapped it with HDI.Our results align with Bano et al. 68 , Çakar et al. 120 , Zhu 121 , and Song et al. 122 , who also found that human capital development is a crucial predictor of maintaining a low-carbon environment.Moreover, the findings also indicate that PGR is strongly significant in impacting CO 2 emissions across all panels.It shows that a 1% increase in PGR causes CO 2 emissions to increase by 0.159, 0.19, 0.379, 0.301, and 0.121 metric tons per capita in the full, high, upper-middle, lower-middle, and low-income panels, respectively, in the long run, while short-run effects are insignificant.The positivity of PGR can be traced through two conduits.First, growth in the population, especially uncontrolled growth, increases the demand for energy consumption, industry, and transportation alike, which significantly contributes to increasing CO 2 .Second, PGR is a significant predictor of increases in greenhouse gas emissions.Studies by Dong et al. 123 , Weber and Sciubba 23 , and Ray and Ray 124 support our findings on the positive impact of PGR on CO 2 emissions.
With respect to urbanization (URB), the findings reveal that while its short-run effects are only evident in the full and low-income panels, its long-run effects are significant across all panels.It shows that URB is another factor that, without exception, increases CO 2 emissions.The findings are linked to the fact that higher urbanization results in greater deforestation, higher freshwater extraction, and the utilization of more carbon-producing goods that reduce environmental quality in the long run 125 .Prior studies by Akalin et al. 126 , Nathaniel 127 , Kahouli et al. 128 , and Radoine et al. 129 also support our findings.Finally, our findings with respect to trade openness (TOP) are somehow similar to the existing literature.We only find that TOP is significant in high-income panels in the short run, while it only affects CO2 emissions in UMIC, LMIC, and LIC panels in the long run.Overall, our findings indicate that TOP would facilitate higher CO2 emissions.Studies that concur with our findings include Ertugrul et al. 130 , Ragoubi and Mighri 131 , Dou et al. 22 , Chen et al. 132 , and Adebayo et al. 108 , though studies by Mahmood et al. 133 and Yu et al. 134 found negative and spillover effects of TOP on CO 2 emissions, respectively.
All results reported in Policy implications.Our findings highlight several policy implications that are specifically discussed as follows: i. Altogether, good governance is crucial to maintaining and improving global environmental quality, regardless of the size of its effects.It is imperative to institutionalize good governance to encourage efficient reiteration and utilization of resources for higher environmental preservation.ii.For high-income countries, economic growth is no longer a silver bullet to recast environmental quality; rather, it is regarded as another essential tool for upper-middle-income, lower-middle-income, and lowincome countries to reverse the negative impact of rapid growth on environmental quality.This suggests specific policy reorientations in their growth-targeting regimes.iii.Although budget implications are real concerns in low-income countries, the findings suggest that efficient energy consumption and the deployment of innovative ecological technologies in the production sector of the economy can spur environmental quality.iv.All in all, the growth and massification of populations in urban areas are harmful to environmental quality.
Specific policy adjustments are required to facilitate the economic shift and ensure an even population distribution.v.With no exception, human interaction with the environment is also a determinantal factor.Well-thoughtout investments in human capital development can result in increased education and awareness to preserve environmental quality.vi.From a global perspective, while many factors contribute to global warming, CO 2 is the most important, implying economies must follow global policy incentives and implement new mechanisms to reduce CO 2 , such as better forest management, taxes on ecologically harmful behaviors, increasing the total cumulative area of the Earth sheltered in forests, and smoothing the transition to electric and hybrid automobiles.
Limitations.This study highlights an important promotional role for the links between good governance and a sustainable environment and has provided a comprehensive statistical scenario of the effects of good governance on environmental quality from a global perspective, but it suffers from one major limitation: the exclusion of armed conflict effects from the analysis in some of the countries due to the unavailability of relevant datasets.Future studies can overcome this empirical shortcoming, depending on the availability of the required data.

Figure 1 .
Figure 1.World CO 2 emissions.Values are shown in natural logarithmic form.Source: Our World in Data.

value Statistics p-value Statistics p-value Statistics p-value Statistics p-value
91vironmental degradation, while it keeps reducing in upper-middle-income countries and even lowers in lower-middle-income and low-income countries.This might be due to the financial, technical, and social commitment of the countries towards the implementation of good governance.The findings are theoretically valid and acceptable.High-income economies, comparatively, have institutionalized good governance as an integral part of their normal administrative endeavors, while a high corruption tendency and lower rule of law are signaled in low-income economies.Our results are partially consistent with the findings of Vogel89, Bhattarai and Hammig 90 , Ehrhardt-Martinez et al.91, Cole and Neumayer 92 , Welsch 93 , Esty and Porter 94 , Fan et al. 95 , Culas 96 , Newell 97 , Berkman and Young 98 , Bulkeley 99 , Arvin and Lew 100 , Pour 101 , Newell et al.
Table 7 are statistically robust.Diagnostic checks are reported underneath every panel estimation of the CS-ARDL model.They report two important facts.First, CD is corrected across all panels.www.nature.com/scientificreports/environment worldwide.It implies that as a result of population growth and higher urbanization, demand for energy, industry, and transportation increases, resulting in increased CO 2 emissions.Comparatively, the results demonstrate that financial development negatively effects CO 2 , implying that a well-developed financial sector facilitates enhanced access to higher investments in lower carbon production that significantly decreases CO 2 emissions.